The Use of Smart Phones to Estimate Road Roughness: A Case Study in

Turkey

Metin Mutlu AYDIN1, Mehmet Sinan YILDIRIM2 and Lars FORSLOF3

1
Akdeniz University, Civil Engineering Department, Antalya, 07158, Turkey,
metinmutluaydin@gmail.com
2
Manisa Celal Bayar University, Civil Engineering Department, Manisa, 45140, Turkey,
msyildirim35@gmail.com
3
Roadroid AB, Ljusdal, Sweden, lars.forslof@roadroid.com

ABSTRACT

Previous studies has shown that road surface conditions are an important factor for road
quality, and smooth roads for providing more comfortable and more safety driving
experience. To provide quality on road surface, it should be observed steadily and repaired as
necessarily. There are many process to determine road surface condition which usually
requires high costs and skillful operators such as visual inspection and instrumented vehicles
that can take physical measurements of the road surface deformations. To estimate road
roughness, Android OS based smart phones that has a mobile sensing system can be used for
road surface condition detection. Using a smart phone to collect data is an alternative and
simple application because of it’s low cost, wider population coverage property and easy
utilization. This paper explores the utilization of Roadroid, a simple android application, as a
low cost vehicle-based solution for road surface condition monitoring with using sensors
from smartphones. In the scope of this study, site experiments have been conducted to collect
data using acceleration and GPS properties of a smartphone in a specific (passenger car)
vehicle type. This method was evaluated with 3259 km urban and rural road data collected
from the site experiments in Turkey, and it was seen from the results that average 84.4% of
Turkish roads have good, 7.9% have satisfactory, 3.8 have unsatisfactory and 3.8% have poor
road roughness conditions. It shows that approximately 4% of Turkish roads need
maintenance urgently. Also experimental study results confirm that Roadroid have a great
potential to evaluate road surface roughness condition correctly, even under obstacle like,
potholes, manholes and decelerating marks.

Keywords: Road surface condition, road roughness, smartphone sensors, mobile sensing,
Roadroid.

I. INTRODUCTION

Road surface conditions are have key importance for the safety movement of vehicles. Road
pavements are generally designed based on an estimated traffic carrying capacity throughout
a determined service life [1,2]. After the pavement surface begins to its service life, various
road surface deformations (such as potholes, broken edges, rutting, cracking, swelling etc.)
occur due to road infrastructure, superstructure deficiencies and excessive heavy vehicle
International Conference on Advanced Engineering Technologies ( ICADET 2017), 21-23 Sept. 2017, Bayburt
loads [2]. Deformations or poor surface conditions may have adversely affect the service
ability of a road in relation with their position and size. Additionally, the deformations cause
persistent problems especially in undeveloped and developing country roads due to low-cost
road policy [3]. Also developed countries have road surface problems caused by climatic
conditions, road work zones and excessive vehicle loads etc. Especially, municipalities all
over the world spend big moneys to maintain and repair their roadways [4]. Therefore,
monitoring road surface conditions has received a significant amount of attention by road
authorities, planners and researchers [5]. In order to determine these mentioned poor surface
conditions problem, many researchers have been working to monitor these deformations by
using special devices (such laser road surface scanner, light detection and ranging, and
mobile phones) and techniques (such as GPS devices and accelerometers in vehicles or
cameras on roadside and near traffic signals) [4-13].
The IRI (International Roughness Index) is commonly used to measure the road roughness as
an indicator to determine the pavement surface quality. Since its first introduction (1986), the
IRI standard has become very famous indicator to evaluate and manage road surface
pavement quality [14]. To measure IRI, many methods are used and most of them require
sophisticated profilers and tools. These profilers and tools are too expensive to acquire and
operate. Additionally, in economically weak countries visual measurements have also
popular utilization. This method is a cheap option but it is usually very labor intensive and
time consuming [14].
Nowadays, using smartphones for monitoring road surface quality data have a very popular
utilization because of its low cost and easy implementation. Smart phones have many useful
sensors inside of them. A 3D or 3-Axis accelerometer is one of the most important sensor in a
smartphone [15]. An accelerometer can measure the acceleration in m/s2 along each axes (x,
y and z). For this reason, smartphones are generally used to determine motion activities. In
literature, there some studies about utilization of smartphones’ accelerometer property to
detect road bumps and anomalies [4,5,10,16]. But smartphones are generally used with
developed road roughness measurement applications especially classifying roughness
condition of road section by using simple techniques. In literature, there are some
applications that can measure the road roughness such as Nericell [5], TrafficSense [17],
Roadroid [18] etc.
Unfortunately, road authorities in Turkey have a not a common method to observe road
surface quality. They generally choose expensive methods to observe surface quality of
Turkish urban and rural roads. Whereas, many urban and rural roads’ surface qualities can be
measured easily and very cheaply with the help of smartphone applications. For this purpose
in this study first time, a road roughness smartphone application was used to determine road
roughness property of Turkish roads. With this method, one of the famous road roughness
measurement application, Roadroid, was used to evaluate the current situation of Turkish
roads and performance of smartphone application on road roughness measurement.

International Conference on Advanced Engineering Technologies ( ICADET 2017), 21-23 Sept. 2017, Bayburt
 II. GENERAL INFORMATIONS ABOUT ROADROID

Roadroid is a Swedish innovation smartphone application and it was developed by Roadroid

Company. It measures road roughness by using the accelerometer and Global Positioning
System (GPS) property of the smartphones [19]. It supports different vehicle types, car
speeds and phone models. In the field trials and calibration process of the application, over 10
million sample data were collected globally. Roadroid also uses Android programming and
global mapping service inside. It is continuously developed and new versions were released
by the developer company. Basically, Roadroid application analyze the vibration data and
estimate an International Road Index (eIRI) value by considering current speed, vehicle type
and used smartphone model (Figure 1). The correlation of the application was originally
developed towards Swedish IRI measurements by using laser beam [18-20].
Roadroid is a Swedish innovation smartphone application and it was developed by Roadroid
Company. It measures road roughness by using the accelerometer and Global Positioning
System (GPS) property of the smartphones [19]. It supports different vehicle types, car
speeds and phone models. In the field trials and calibration process of the application, over 10
million sample data were collected globally. Roadroid also uses Android programming and
global mapping service inside. It is continuously developed and new versions were released
by the developer company. Basically, Roadroid application analyze the vibration data and
estimate an International Road Index (eIRI) value by considering current speed, vehicle type
and used smartphone model (Figure 1). The correlation of the application was originally
developed towards Swedish IRI measurements by using laser beam [18-20].

(a)

International Conference on Advanced Engineering Technologies ( ICADET 2017), 21-23 Sept. 2017, Bayburt
 (b)
Figure 1. A General Overview to Roadroid Application [21].

Roadroid has easy usable, user friendly and cost efficient probability to measure road
roughness in different road, vehicle, speed, weather and smartphone conditions. It saves IRI
values with longitudinal and lateral positions, altitude and vehicle speed each second, and
obtained data is presented on an internet GIS tool. User can extract data with 100 meter
section. Also in the process of data extraction roughness data, vertical road profile and
vehicle speed can be mapped. Obtained results of the system have shown that the correlation
of the estimated IRI (eIRI) towards laser beam measured IRI is about 70-80% depending on
road surface types [21]. The accuracy of the system can be increased with some tuning and
the IRI sampling is currently developed with a calculated IRI (cIRI) to enhance the
correlation factor.
Roadroid application use also the camera property of the smartphones. It can take GPS
tagged photos from the site in the data collection process and it can be transferred to the map
tool of the system (Figure 2).

(a) (b)
Figure 2. Site Observation Photos from Turkish Roads Taken By the Roadroid Application.

This application is mainly supports early warning system to road user and road maintenance
authorities. For example, in winter season daily road surface data can be shared by the grader
International Conference on Advanced Engineering Technologies ( ICADET 2017), 21-23 Sept. 2017, Bayburt
operations and grader operators can make a plan to clean roads. Also in winter and spring
seasons, roads are deformed easily caused by the frost heave effect. Road users or road
maintenance authorities can monitor the roughness of the roads day by day. Then they can
clearly see when and where the problem occurs on road surfaces

III. SITE OBSERVATION AND DATA COLLECTION

In this study, Researchers from Akdeniz and Manisa Celal Bayar Universities have worked as
Turkey Partner of Roadroid Company. In the site observations a Samsung Galaxy S2
smartphone (supplied by Roadroid Co.) with Roadroid Classic Application has been used to
observe and measure road roughness (Figure 3). For this purpose, only passenger car vehicle
type is used in data collection process and vehicles speeds were chosen between 0-100 km/hr.
But Roadroid application can only calculate eIRI when driving speed of experiment vehicle is
20 km/h or faster [21]. For this reason, some data were not obtained because of the low speed
(because of the interrupted traffic flow effects such intersection effect, high traffic volume
effect, signal effect etc.).

(a) (b)
Figure 3. Used Galaxy S2 Smartphone with Roadroid Road Roughness Measurement Application.

To make a site observation, firstly system should be installed according to given flowchart as
given below (Figure 4).

Figure 4. Installation of Roadroid System into the Car.

International Conference on Advanced Engineering Technologies ( ICADET 2017), 21-23 Sept. 2017, Bayburt
Measurements were conducted in different urban and rural roads of 18 various cities in
Turkey (Table 1). It means approximately 22% (18 cities) city roads of (81 cities) total cities
were used as example for the analysis. All research data were collected between 2014-2017
years and shared with Roadroid Co. as Turkey Partner. Total 3259 km urban and rural roads
were observed during this period. From the previous studies, it was found that different road
surface conditions cause vehicles to vibrate differently [15]. For this reason in this study,
placing smartphones that come with acceleration sensors, the variation of the vibration is
believed to be captured as become Douangphachanh and Oneyama [15]’s study. Then drivers
drove the vehicles with normal driving conditions along many roads that have different
surface and traffic conditions.
Table 1. Observed Cities of Turkish Roads.
City Name - (Plate No City Name (Plate
No
Number) Number)
1 Ankara - (06) 10 Eskişehir – (26)
2 Afyonkarahisar –(03) 11 Gümüşhane – (29)
3 Amasya – (05) 12 Isparta – (32)
4 Antalya – (07) 13 Kütahya – (43)
5 Artvin – (08) 14 Rize – (53)
6 Burdur – (15) 15 Samsun – (55)
7 Bursa - (16) 16 Trabzon – (61)
8 Çorum – (19) 17 Bayburt – (69)
9 Erzurum – (25) 18 Kırıkkale – (71)
After the site observations, data were uploaded to Roadroid website and all collected data
were obtained from the Roadroid website data import system as shown in Figure 5.

Figure 5. Data Obtaining From the Roadroid Website.

After obtaining data from the Roadroid website, all data were extracted to the excel
spreadsheets. Then all spreadsheets are carefully controlled to determine estimated
International Road Index (eIRI) data. All data were obtained from the system by selecting the
aggregation lengths (meters) as 100 meters. It was seen from the data extraction process that
sections with incomplete data set are the sections that have no data from Roadroid. Because,
Roadroid cannot calculate an eIRI value when vehicle speed < 20 km/hr as mentioned
previously.
International Conference on Advanced Engineering Technologies ( ICADET 2017), 21-23 Sept. 2017, Bayburt
 IV. ANALYSIS AND FINDINGS

In this study, during the observation period total 3259 km different urban and rural roads of
Turkish cities were measured by using Roadroid road roughness measurement application
(Figure 6).

Figure 6. Example Data Analysis from Different Cities of Turkey.

All obtained results for these 18 city roads are given in Table 2. As can be seen from the
Table 2, Eskişehir has the highest and Burdur has the lowest road roughness properties.
According to analysis results, five cities’ (Eskişehir, Isparta, Bursa, Çorum and Samsun
cities, respectively) good road roughness properties are above 90%. On the other hand, two
cities’ (Burdur and Artvin) good road roughness properties are below 70%. It can be
concluded that Burdur and Artvin roads need maintenance earlier than the other 16 cities.
Also according to general results, average 84.4% of Turkish roads have good, 7.9%
satisfactory, 3.8 unsatisfactory and 3.8% poor road roughness conditions. It shows that
approximately 4% of Turkish roads need maintenance urgently.
eIRI values obtained from the analysis provides a reasonable assessment of the pavement
condition, which can provide useful inputs for pavement maintenance decision making.
According to site observation and analysis results, it can be concluded that eIRI is estimated
for the gravel roads which have very poor surface conditions. From the literature, it can be
seen that the minimum acceptable IRI values vary in different countries [22]. It effected by
the maintenance type and plan, road type and budgetary of road maintenance authorities. For
example, an acceptable IRI threshold value would be 2 due to higher operating speeds
expected. FHWA criteria acceptable for arterial roads is about 3.5 [23]. However this value
can be changed for low volume roads with lower operating speeds. Typically, for low volume
roads, pavement with IRI in the range of 6-10 is considered to be in moderate/poor condition
[23]. Once these threshold values are established by the road authorities, they first select the
roads that violate the threshold and then prioritize the roads for maintenance works for those
roads based on the condition and the maintenance treatment plan.

International Conference on Advanced Engineering Technologies ( ICADET 2017), 21-23 Sept. 2017, Bayburt
 Table 2. Analysis of Road Roughness of Various Cities in Turkey.
Good Satisfactory Unsatisfactory Poor Mean Total
City Name - (Plate
No Point Point Point Percentage Point Value Number
Number) Percentage (%) Percentage (%) Percentage (%)
Number Number Number (%) Number (eIRI) of Points
1 Ankara - (06) 6342 88.8 460 6.4 187 2.6 153 2.1 1.84 7142
2 Afyonkarahisar –(03) 5657 89.2 367 5.8 120 1.9 198 3.1 1.92 6342
3 Amasya – (05) 4169 85.6 430 8.8 153 3.1 121 2.5 1.94 4873
4 Antalya – (07) 33417 81.2 3815 9.3 1837 4.5 2109 5.1 2.30 41178
5 Artvin – (08) 5250 69.9 759 10.1 392 5.2 1109 14.8 3.04 7510
6 Burdur – (15) 5285 62.5 1742 20.6 788 9.3 637 7.5 2.92 8452
7 Bursa - (16) 4208 94.9 94 2.1 59 1.3 72 1.6 1.17 4433
8 Çorum – (19) 3824 92.7 223 5.4 37 0.9 41 1.0 1.42 4125
9 Erzurum – (25) 3241 83.2 308 7.9 199 5.1 148 3.8 2.19 3895
10 Eskişehir – (26) 2149 98.1 25 1.1 11 0.5 5 0.2 1.38 2190
11 Gümüşhane – (29) 11353 76.0 1864 12.5 767 5.1 947 6.3 2.57 14931
12 Isparta – (32) 11435 95.1 356 3.0 154 1.3 84 0.7 1.48 12029
13 Kütahya – (43) 7067 72.1 1620 16.5 704 7.2 411 4.2 2.28 9802
14 Rize – (53) 3964 86.3 294 6.4 179 3.9 156 3.4 1.95 4593
15 Samsun – (55) 5023 90.6 172 3.1 143 2.6 207 3.7 1.70 5545
16 Trabzon – (61) 2493 80.2 290 9.3 203 6.5 123 4.0 1.95 3109
17 Bayburt – (69) 2050 88.3 156 6.7 71 3.1 45 1.9 1.80 2322
18 Kırıkkale – (71) 3088 84.8 288 7.9 157 4.3 109 3.0 2.01 3642
Average (μ) 6667 84.4 737 7.9 342 3.8 371 3.8 1.99 8117

International Conference on Advanced Engineering Technologies ( ICADET 2017), 21-23 Sept. 2017, Bayburt
 V. CONCLUSION AND DISCUSSION

Previous studies has shown that road surface conditions are an important factor for road
quality, and smooth roads will provide more comfortable and more safety driving experience.
To provide quality on road surface, it should be observed steadily and repaired as necessary.
There are many process to determine road surface condition which usually requires high costs
and skillful operators such as visual inspection and instrumented vehicles that can take
physical measurements of the road surface deformations. To estimate road roughness,
Android OS based smart phones that has a mobile sensing system can be used for road
surface condition detection. Using a smart phone to collect data is an alternative and simple
application because of it’s low cost, wider population coverage property and easy utilization.
This paper explores the use of Roadroid, one of the famous road roughness measurement
smartphone application, as a low cost vehicle-based solution for road surface condition
monitoring with using sensors from smartphones.
In the scope of this study site experiments have been conducted to collect data using
acceleration and GPS properties of a smartphone in a specific (passenger car) vehicle type.
The application is used on 18 different cities and total 3259 km urban and rural roads data in
Turkey, and it was seen from the results that average 84.4% of Turkish roads have good,
7.9% satisfactory, 3.8 unsatisfactory and 3.8% poor road roughness conditions. It shows that
approximately 4% of Turkish roads need maintenance urgently. Also experimental study
results confirm that Roadroid have a great potential to evaluate road surface roughness
condition correctly, even under obstacle like, potholes, manholes and decelerating marks.
Mainly, this study was a pilot study to see Roadroid application’s performance on Turkish
roads. For this reason, it is applied only 18 city and total 3259 km roads in Turkey. In this
study, urban and rural roads are evaluated with together because of the limited urban road
data. Therefore, in future studies, total 81 city roads should be examined and data should be
evaluated for urban and rural separately.

References
[1] J. Ben-Edigbe, N. Ferguson, “Extent of capacity loss resulting from pavement distress,” Proceedings of
the Institution of Civil Engineers – Transport, vol. 158, no. 1, pp. 27–32, 2005.
[2] M.M. Aydin and A. Topal, "Effect of road surface deformations on lateral lane utilization and
longitudinal driving behaviours", TRANSPORT, vol. 31, no. 2, pp.192-201, 2016.
[3] J. Ben-Edigbe, “Assessment of Speed–Flow–Density Functions under Adverse Pavement Condition”,
International Journal of Sustainable Development and Planning vol. 5, no. 3, pp. 238–252, 2010.
[4] J. Eriksson, L. Girod, B. Hull, R. Newton, S. Madden and H. Balakrishnan, “The pothole patrol: using a
mobile sensor network for road surface monitoring,” In Proceedings of the 6th International Conference
on Mobile Systems, Applications, and Services, pp. 29-39, 2008.
[5] P. Mohan, V.N. Padmanabhan and R. Ramjee, “Nericell: rich monitoring of road and traffic conditions
using mobile smartphones,” In Proceedings of the 6th ACM Conference on Embedded Network Sensor
Systems, pp. 323-336, 2008.
[6] TRB, “Automated Pavements Distress Collection Techniques: a Synthesis of Highway Practice”.
NCHRP Synthesis 334, National Cooperative Highway Research Program (NCHRP). Transportation
Research Board (TRB), Washington, DC. 94 p., 2004.

International Conference on Advanced Engineering Technologies ( ICADET 2017), 21-23 Sept. 2017, Bayburt
[7] A. Oloufa, H. Mahgoub and H. Ali, “Infrared Thermography for asphalt crack imaging and automated
detection”, Transportation Research Record: Journal of the Transportation Research Board, vol. 1889,
pp.126–133, 2004.
[8] H.D. Lee and J.J. Kim, “Development of a Manual Crack Quantification and Automated Crack
Measurement System”. Project TR-457 Final Report. University of Iowa, US. 21 p., 2005.
[9] S. Battiato, L. Rizzo, F. Stanco, S. Cafiso and A. Di Graziano, “Pavement surface distress by using non-
linear image analysis techniques,” in Proceedings of SIMAI 2006, pp. 1-4, 2006.
[10] G. Strazdins, A. Mednis, G. Kanonirs, R. Zviedris and L. Selavo, “Towards vehicular sensor networks
with android smartphones for road surface monitoring”, 2nd international workshop on networks of
cooperating objects, 2011.
[11] V. Bychkovsky, K. Chen, M. Goraczko, H. Hu, B. Hull, A. Miu, E. Shih, Y. Zhang, H. Balakrishnan,
and S. Madden, “The cartel mobile sensor computing system,” In SenSys’06, pp. 383–384, 2006.
[12] J. Yoon, B. Noble, and M. Liu, “Surface street traffic estimation,” In MobiSys 07, pp. 220–232, 2007.
[13] R. Sen, B. Raman, and P. Sharma, “Horn-ok-please,” In MobiSys, pp. 137–150, 2010.
[14] M.W. Sayers and S. Karamihas, “Interpretation of Road Roughness Profile Data,” FHWA/RD-96/101
University of Michigan. 1996.
[15] V. Douangphachanh and H. Oneyama, “Estimation of road roughness condition from smartphones under
realistic settings,” In: ITS Telecommunications (ITST), 2013 13th International Conference on. IEEE, pp.
433-439, 2013.
[16] A. González, E.J. O’brien, Y.Y. Lia and K. Cashell, “The use of vehicle acceleration measurements to
estimate road roughness”, Vehicle System Dynamics, vol. 46, no. 6, pp. 483–499, 2008.
[17] Trafficsense, “Rich monitoring of road and traffic conditions using mobile smartphones,” Microsoft
Research, Tech. Rep. MSR-TR-2008- 59, 2008.
[18] L. Forslöf, “Roadroid–smartphone road quality monitoring,” In Proceedings of the 19th ITS World
Congress, 2012.
[19] L. Forslöf and H. Jones, “Roadroid: continuous road condition monitoring with smart phones,” In IRF
17th World Meeting and Exhibition, 2013.
[20] L. Forslöf and H. Jones, “Roadroid: Continous Road Condition Monitoring with Smart Phones”, Journal
of Civil Engineering and Architecture, vol. 9, no. 2015, pp. 485-496, 2015.
[21] Roadroid Guideline, Quick start ver 1.2.1., 2013, Sweden.
[22] R.O. Souza, “Influence of Longitudinal Roughness on the Evaluation of Pavement,” M.S. Dissertation,
Publication 625.8(043) S729i, University of Brasilia, Brasilia, Brazil. (In Portuguese), 2002.
[23] D. Gamage, H.R. Pasindu and S. Bandara, “Pavement Roughness Evaluation Method for Low Volume
Roads,” Proc. of the Eighth Intl. Conf. on Maintenance and Rehabilitation of Pavements, 2016.

International Conference on Advanced Engineering Technologies ( ICADET 2017), 21-23 Sept. 2017, Bayburt